Description of the

Small Parts Demo Model

for Tecnomatix Plant Simulation
November 2011

CONTENT
1. Description ............................................................................................ 2
1.1 Objective .................................................................................................. 2
2. Demo Instructions................................................................................. 2
2.1 Study Overview ......................................................................................... 2
2.2 System Parameters ................................................................................... 3
2.3 Find the Optimum number of Pallets (*) ....................................................... 5
2.4 Find the Bottleneck ................................................................................... 8
2.5 Test New Layout ....................................................................................... 9
3. Worker Utilization ............................................................................... 10
4. Conclusion ........................................................................................... 10

Small Parts Production Demo Model Description

1. Description
This model shows a production and assembly system with a pallet-based transport system. The system
contains manual and automatic Workstations. One part per pallet runs through the system and is
processed on the stations according to the process times.
Each station has cycle times, certain availability and the manual stations need a worker to start. These
parameters determine the time a part stays on the station.
Note, that some actions in this description are not available in the Plant Simulation Viewer (in case you run
this model from the Plant Simulation Demo-CD), since the Plant Simulation Viewer does not allow any
modeling or changes of parameters. Actions which are not allowed in the Plant Simulation Viewer are
marked with an asterix (*). In the model for the Plant Simulation Viewer, some of the actions marked with
an asterix (*) will happen automatically at the end of the simulation.

1.1 Objective
The objective is to optimize the number of workers, pallets and the capacity of buffers to maximize the
throughput. Therefore, this model shows how to solve two typical problems every simulation engineer has:
1. In a real production system, there are always several values to optimize (e.g. maximize throughput
rate and minimize throughput time) and several parameters you can change (capacities, logic,
layout). In most cases, some parameters affect other parameters, too (e.g., if you decide to
decrease the capacity of buffers, the optimum number of pallets might decrease, too). This results
in the problem, that a user should run experiments for all possible combinations of values for
those parameters to find the optimal result result. Normally, this is not possible. The
ExperimentManager in Plant Simulation can solve this problem by executing a certain number of
simulation runs by itself and by using neural networks to calculate the optimum combination of
parameters from the results.
2. Some values in a real system have a random behavior, e.g. the point in time when a machine
breaks down or the cycle time of manual operations. In this case, its not sufficient to run just one
simulation experiment, because the results of this simulation experiment are based on the random
numbers from this experiment. When you run the same experiment with different random
numbers, you will get different results. So how can you base a decision over a multi-million dollar
system on simulation experiments? The solution is, to run multiple experiments with different
random numbers and to calculate a confidence interval from the results. The ExperimentManager
in Plant Simulation can tell you that e.g. with a probability of 99 % (which is the confidence
level), the mean throughput of your system will be in the range of 45.2 to 45.5 units per hour.

2. Demo Instructions
2.1 Study Overview
Start a simulation run in the frame Assembly1 using the button Start in the model.

Small Parts Production Demo Model Description

Figure 1: Start/Stop icon in the toolbar

Watch the simulation run. On the left hand side, pallets enter the system. The LoadStation on top loads
parts on the pallet. Then, the pallet moves through several manual and automated workstations. When the
pallet enters a manual workstation, a worker is allocated from the worker pool. At station MS3 on the right
hand side, assembly parts are assembled on the main part. The assembly parts arrive from the station
PreProduction. Double click this station to look at the content.
Then, the pallets with the main part move on to additional stations. At the bottom, 40 % of the parts are
loaded on a cross transfer element and have to pass a test station. On the UnloadStation, the main part is
removed from the pallet and leaves the system. The pallet moves on to take the next part.

Figure 2: The model Assembly1

2.2 System Parameters

Right-click the attribute explorer ManuParameters and choose Show from the context menu. You can see
that we use the Normal distribution to create random numbers for the cycle time. All stations have an
Availability of 98 %.

Small Parts Production Demo Model Description

Figure 3: Attribute Explorer for the manual workstations

Right-click the attribute explorer AutoParameters and choose Show from the context menu. You can see
that we use the Uniform distribution to create random numbers for the cycle time. All stations have an
Availability of 99 %.

Figure 4: Attribute Explorer for the automated workstations

To change the number of pallets (*), double-click the Quantity object and change the value in the dialog.

Figure 5: Quantity of pallets

We use a 2-shift system in this model. Double-click the ShiftCalendar to look at the shift times.

Small Parts Production Demo Model Description

Figure 6: Shift calendar

To change the number of workers (*), double click the Pool object and click on Creation Table. In the table
which opens, you can see the type worker we use and the quantity for each shift. The column Worker
refers to the type of worker you have in the class library (different workers can do different jobs).

Figure 7: Creation table of the worker pool

Note that the shift name in the column Shift corresponds to the shift name in the ShiftCalendar.

2.3 Find the Optimum number of Pallets (*)

We would like to find out, what is the right number of pallets to maximize the throughput of the system.
Besides that, we would like to run several simulation experiments with different random numbers, to get
statistically reliable results.
Double-click the ExperimentManager object. On the tab Definition click on Define Output Values. This
table shows the result value we would like to optimize. Then click on Define Input Variables. This table
5

Small Parts Production Demo Model Description

shows the parameter we would like to change to optimize the result. In both tables you can provide
understandable descriptions that are used in the evaluation of the experiments.
A click on Define Experiments opens a table which shows in which values we would like to set. On the tab
Definition, you find the number of simulation runs with different random numbers (Observations).

Figure 8: The ExperimentManager tab Definition

Click on the bottoms Reset and Start. Now the ExperimentManager executes 5 simulation runs
(5 observations) for each input value we defined. We defined 11 steps, so we execute 5 * 11 = 55
simulation runs. Each run simulates 24 hours. At the end, an HTML report opens.

Figure 9: Overview over all executed expriments

6

Small Parts Production Demo Model Description

Look at the page Overview of the report. It shows the input and output values for each experiment. On
the page Output values of interest, you see a chart which shows the confidence interval for each step. It
shows for each experiment the range in which the mean value of the throughput will be with a 90 %
probability.

Figure 10: The confidence intervals of the throughput

From the HTML report we see, that experiment 7 shows the highest throughput. In experiment 7 we used
27 pallets.
Double-click the Quantity object in the Assembly1 frame and set the value to 27. Click on the Reset
button and on the Start button of Assembly1.
Summary: Using the ExperimentManager of Plant Simulation, you can very easily run multiple
simulation experiments with different random numbers to verify if your simulation results are
reliable.

Small Parts Production Demo Model Description

2.4 Find the Bottleneck

Right-click on the Utilization chart and choose Show from the context menu show the utilization of the
stations.

Figure 11: Utilization chart of the stations

The chart shows the percentage of time where a station was

Working (green)

Waiting (grey)

Blocked (yellow)

Failed (red)

Stopped (rose)

Paused (dark blue)

Unplanned (light blue)

From the chart you can see, that the stations MS3, MS4 as well as AS2 and AS3 show a certain blocking
percentage (yellow). On the other hand side, you see that most stations show a waiting percentage (grey).
This indicates that there is a bottleneck at stations AS3/AS4 and MS4.
Right-click the Buffer Usage histogram shows the percentage of time a certain quantity of parts was located
on a conveyor.

Small Parts Production Demo Model Description

Figure 12: Occupation of conveyors

Note, that the conveyors C4 and C5 (yellow and turquoise) are quite often fully occupied. Again, this
indicates that there is a bottleneck at station AS3/MS4. So the problem is the direct connection of manual
and automated workstations. To solve the problem, we insert a buffer B between MS4 and AS3.

2.5 Test New Layout

Look at the frame Assembly2. It shows a buffer between MS4 and AS3. Double-click the buffer. In the
dialog of the buffer you see, the buffer has a capacity of 6 parts. Close the dialog.
Open the ExperimentManager in Assembly2 and run the experiment study again (*). Look at the Output
values of interest of the HTML report.

Figure 13: The confidence intervals of the throughput of Assembly2

In the report you see, that the throughput of Assembly2 is higher, but the optimum number of pallets now
is 29 pallets. The experiment 8 shows the highest throughput. So the insertion of the buffer increased the
number of required pallets. The mean throughput of the produced parts can be improved from 1376.4
(Experiment 7 with 27 palettes in model Assembly1) to 1410.2 (Experiment 8 with 29 palettes in model
Assembly2).
9

Small Parts Production Demo Model Description

Double-click the Quantity object in the Assembly2 frame and set the value to 29. Click on the Reset button
and on the Start button in the toolbar of Assembly2. Open the utilization chart for the stations and
compare the result.

3. Worker Utilization
To look at the utilization of the workers, right-click the Worker-Utilization object and choose View Chart
from the context menu.

Figure 14: The worker utilization chart

The diagram shows the percentage of time when an operator was

Working (green)

Waiting (grey)

Paused (dark blue)

Unplanned, meaning there was no shift time (light blue)

The chart tells you that the workers at the manual workstations have a high utilization. The utilization of
the Adjusters is very low.

4. Conclusion
Even a system that looks so simple shows a complex behavior, due to stochastical parameters.
Even if two stations have the same mean value of the cycle time, the second station can be a bottleneck, if
the two stations use a different random numbers distribution.
Summary: Plant Simulation provides all the tools you need to base your decisions on reliable
results. Plant Simulation provides multiple easy-to-use tools to evaluate your system. These tools
allow you to identify bottlenecks and find resources which are not well utilized.

10

Small Parts Production Demo Model Description

2008 Siemens Product Lifecycle Management Software II (DE) GmbH. All Rights Reserved.
This documentation is proprietary to Siemens Product Lifecycle Management Software II (DE) GmbH.
This document contains proprietary information and is protected by copyright. No part of this document
may be reproduced, stored in a retrieval system, translated, transcribed, or transmitted, in any form or by
any means, without the prior explicit written consent of Siemens Product Lifecycle Management Software
II (DE) GmbH.
Siemens and the Siemens logo are registered trademarks of Siemens AG. Tecnomatix and the
Tecnomatix logo are registered trademarks of Siemens Product Lifecycle Management Software Inc.
All other product names or brand names are trademarks or registered trademarks of their respective
owners.
Information in this document is subject to change without notice.

11